我国极端温度事件的中期天气可预报性和集合概率预报方法研究

In recent years, the extreme high- and low-temperature events occurred frequently in China, usually accompanied with huge societal influences and economical losses, however, which forecasting only has a quite limited performance until now. Researches on the medium-range weather forecast of extreme temperature event (ETE) in China are urgently needed as the short-range forecast becomes more and more difficult to satisfy the increasing public requests. This project aims to investigate the 6‒15d medium-range weather predictability and develop the ensemble-based probabilistic forecasting methods for the ETE in China. First, the climatic percentile approach is used to identify the ETEs in China and then the features and mechanisms of the typical ETEs and their corresponding atmospheric circulation anomaly patterns are examined. Secondly, utilizing a large sample size of reforecast and forecast datasets from ensemble forecasting of numerical models, the performance of the model ensemble in forecasting the China ETEs is verified and evaluated. Thirdly, based on ensemble forecast information, the medium-range predictability is deeply studied for the winter extreme low-temperature event and summer extreme high-temperature event. Finally, by introducing the nonlinear information of model-predicted near-term evolutions and their historical analogues, the new ensemble-based probabilistic forecasting methods, suitable to the 6‒15d medium-range weather forecast of the China ETEs, are developed and further applied to conducting forecast experiments of the ETE based on probabilistic ensemble information. Through this project, it will be expected to add understanding of the 6‒15d medium-range weather predictability the ETE in China and then provide a valuable reference for improvement of the medium-range weather forecast of the ETE.

近些年来，我国极端高温/低温天气事件频发，社会影响和经济损失巨大，但目前预报能力仍很有限，短期天气预报愈发难以满足日益增长的公众需求，亟待开展极端温度事件的中期预报研究。本项目着眼于我国极端温度事件的6-15天中期预报问题，拟深入开展其可预报性和集合概率预报方法研究。首先运用气候百分位方法识别观测中的我国极端温度事件、并考察典型事件自身和环流型等特征及机理。然后利用大样本模式集合预报历史和实时数据，开展集合对极端温度事件预报能力的系统性评估，并重点针对我国冬季极端低温和夏季极端高温事件开展基于模式集合信息的可预报性研究。最后通过引入模式预报的近期演变及其历史相似的非线性信息，发展适用于我国极端温度事件6-15天中期预报的集合概率预报新方法，并开展基于集合概率信息的极端温度事件的预报应用试验研究。通过本项研究，有望提升对我国极端温度事件中期可预报性的认知水平、为改进极端事件中期预报提供参考。

近些年来，我国极端高温/低温天气事件频发，社会影响和经济损失巨大，但目前预报能力仍很有限，短期天气预报愈发难以满足日益增长的公众需求，亟待开展极端温度事件的中期预报研究。本项目着眼于我国极端温度事件的6~15天中期预报问题，深入开展l了其可预报性和集合概率预报方法研究。首先运用气候百分位方法识别观测中的我国极端温度事件、并考察典型事件自身和环流型等特征及机理。然后利用大样本模式集合预报历史和实时数据，开展集合对极端温度事件预报能力的系统性评估，并重点针对我国冬季极端低温和夏季极端高温事件开展基于模式集合信息的可预报性研究。最后通过引入模式预报的近期演变及其历史相似的非线性信息，发展了适用于我国极端温度事件6~15天中期预报的集合概率预报新方法，并开展基于集合概率信息的极端温度事件的预报应用试验研究。通过本项研究，提升了对我国极端温度事件中期可预报性的认知水平、为改进极端事件中期预报提供参考。